Abstract
This research first describes historical perspectives of Least Absolute Value (LAV) estimation, which has been long considered as an estimation alternative of conventional Least Squares (LS) regression. Then, this article explores statistical properties regarding the LAV estimation from Goal Programming (GP). Using a small illustrative example, this study presents new theoretical features regarding the LAV estimation. It is hoped that this research effort enhances its applicability to deal with many decisinal issues in reality.
